Coral

Language learning

Language learning is hard. Here's what has actually helped.

Becoming trulybilingual has been a lifelong goal of mine. Inevitably, it's also one of those aspirational things that gets lost behind other plans (or, honestly, distraction and laziness).

One of my very first attempts at a TODO list somehow included gloriously idiotic line-items like “Learn Mandarin.” How casual my past self was, planning for my future self. Here are some things we've picked up through our own attempts at learning languages, and while digging through every other app in this space.

Immersion

There is truly nothing like being immersed in a language. Ideally, you would live in a country that speaks it: forced to use it to do anything at all, casually overhearing native speakers all day. That isn't always possible. A classroom in a foreign country can help build that kind of environment, or, if you're really disciplined, you can try to surround yourself with the language through movies, shows, podcasts, and music.

Classrooms vs. apps

Language is emotional. Apps can provide good touch points for learning, exposure to the language, and specific lessons, but they can't really replace the emotional context of a classroom. You fit a class into your routine. It's tangible. You think about it. It isn't just sitting somewhere inside your universal computing machine, stripped of context.

You also talk to real people, and practice non-verbal communication. You surround yourself with people worse than you, people at your level, and people better than you. A real person shakes their head when you don't do your homework, too.

Curated, personalized content

Vocabulary lists are great, but they're so often stale. Language is a tool for thought: it helps express what we think and feel. Anchoring your learning to things you actually care about, or thoughts you actually think, is a far better way to build vocabulary than memorizing word lists.

An old colleague once sent me a famous Korean meme built around the word 실패(silpae, roughly “failure”). It isn't the most practical vocabulary in the world, but it's great. I still love saying it, and I'll never forget those words. Making language you care about makes it feel alive, which is exactly how it should be.

Translating words vs. thoughts

Words are just attempts to describe the thoughts in our heads. In programming terms, they're pointers to ideas. The first step of learning a language is usually translating one word or phrase into another word or phrase. The second, and more important, step is to build a mental image from a word or phrase, and then translate that.

I find this to be the root of the issue whenever someone talks about translating something too literally, and it's why great translators are so ridiculously talented.

Spaced repetition

This technique matters, but it's a bit confusing, and maybe a little controversial too. The basic idea is that you don't need to review something every day to remember it. That's inefficient and doesn't scale. There are spacings of review that let you remember more things overall, because you review each individual thing less often.

The controversy tends to come from evangelists who take it too far, or from people who pass around 1,000-card decks and tell you all you have to do is study them. That might be effective, and even necessary, for med school or organic chemistry. It isn't necessary for language. Pacing yourself and building curated, personalized cards is a great way to avoid burnout and the pressure to cram.

Little techniques make a big difference here:

  • Create your own cards whenever you can.
  • Read, speak, or process each card as you make it.
  • When studying, really try to connect with the words, not just translate them.
  • Study sessions should be challenging.

Massive input

There really is no shortcut. There are some annoyingly over-simplified ideas floating around about how children learn language, alongside genuinely interesting studies on the topic. I've even heard the funny counterpoint that kids learn full-time, fully immersed, with heavily-invested personal tutors, and still aren't great by age ten. I don't know the truth of it, but either way, one thing is constant: they get massive amounts of input.

You simply need a lot of time, a lot of effort, or both, if you actually want to learn a language. Games will help you casually pick up a few phrases, and that's great. But if you're genuinely invested, you need more.

Reading while listening

I almost never remember words or phrases unless I see them written down, or write and type them myself. Listening on its own does help: you pick up the overall cadence of the language, and remember a few iconic phrases. It's very likely, though, that you're missing a lot of important subtlety, and that your learning won't compound. By reading words while you listen, you anchor them in your mind and start building the connection between each syllable. Better memory, and a better ability to stretch what you already know into situations you haven't seen yet.

Grammar?!

This includes grammar. As kids, we don't explicitly learn grammar, but that could be down to how much time we spend learning, not some innate language-learning genius.

As adults, we have the disadvantage of being busier, more distracted, and a bit more set in our ways. We have the advantage of being able to understand higher-level concepts, and to use them to learn faster. Grammar helps bridge the gaps in your learning, and clarifies subtleties that would otherwise take a very long time to notice on your own.

The “learning zone”

Balancing novelty and ease is essential to any kind of learning. You have to mix easy and difficult, old and new. Finding where that balance sits for you is difficult, and it's exactly where guided, human learning really shines. It's easy to get demotivated if you're always grinding difficult material for hours at a time. It's just as easy to stagnate if you're only watching shows without really engaging with them or processing them.

There's no shortcut in any of this, and no app replaces the human parts. Coral just tries to be a genuinely useful set of tools for the parts it can help with.